After Seattle-based Amazon began soliciting proposals from North American cities in early September, the business news media was abuzz with speculation about where Amazon will locate its second corporate headquarters. The site will be announced next year.

Of course, not just any city will do. Given the criteria Amazon has laid out, only about 20 metro areas in the United States and Canada are serious contenders. Predictably, though, commentators who fail to see below the surface say that the contest to lure Amazon to one of those cities rather than another will be “transformational” for the winner by creating as many as 50,000 new, high-paying jobs, stimulating economic growth and generating additional tax revenue.

And so, in addition to efforts aimed at convincing Amazon that their city fulfills the company’s requirements, local officials predictably will top up their bids with tax breaks, taxpayer-financed infrastructure upgrades, and similar concessions known in the economic development literature as “selective incentives.” A subsidy by that or any other name would smell as sweet to Amazon’s owners.

If history serves as a guide, politicians will defend giveaways of taxpayer dollars to Amazon by claiming that the benefits flowing from new jobs and higher wages exceed the costs of financing a subsidy. The same arguments routinely are heard when it comes to building a new sports venue, hosting the Olympic Games or SuperBowl, and enticing other headline-grabbing businesses looking to move to greener pastures. Luring Amazon indeed will be a major publicity coup for local and state politicians who certainly will claim credit for a successful outcome in the runup to Election Day.

It turns, out, however, that the benefits of taxpayer-financed subsidies always are overstated. The economic costs of subsidy packages for private business enterprises are in reality much larger than their supporters admit, for several reasons.

First, the contest for Amazon’s second headquarters is a zero-sum game that only one metro area will win. Amazon already has decided to expand and the company will locate somewhere even without a targeted subsidy. Cities competing for Amazon’s new headquarters simply are seeking to shift already planned economic activity geographically, but are not creating any new economic activity.

Second, many economic development studies erroneously count every person hired by a new employer as a job added to the local economy and the new company’s total payroll as an addition to the local tax base. The mistake is that many jobs at Amazon’s second headquarters will be filled by individuals who transfer from existing local jobs (called a “displacement effect”). The taxable income base expands only to the extent that the wages of currently employed workers rise after changing employers. The full wages of people who move to the local area are gains for the city that lures Amazon, but are losses to the places from which they emigrate.

Third, politicians fail to account for the opportunity cost of business-location subsidies. The opportunity cost of any taxpayer-funded subsidy is the private-sector economic activity that would have been generated (but is lost) had the dollars financing it remained in private hands. Nor do economic development studies typically recognize that additional public goods and services (such as public schools, transit systems and police protection)—and additional taxes to pay for them—will be needed to accommodate Amazon’s 50,000 employees and their families.

Finally, politicians argue that new wage earners will spur economic growth through a multiplier effect as their larger incomes circulate around the regional economy. That is certainly true (although multiplier estimates are wildly overstated), but removing income from the private sector to finance subsidies means that the multiplier effect works in reverse. Politicians baldly assume that every dollar of subsidy is worth more to the economy than if the dollar remained in taxpayers’ pockets. Politicians, in other words, believe that they know better than ordinary people how best to allocate resources to their highest valued uses. The historical evidence against that belief is overwhelming.

We won’t know how much the contending urban areas will spend in total to lure Amazon’s second headquarters until the contest is over. New Jersey’s Economic Development Authority and the City of Newark alone have offered an incentive package worth a stunning $7 billion over 10 years. Amazingly, Amazon has received 238 proposals thus far. The social waste in time, money, and effort devoted to buying Amazon’s favor will be monumental, dwarfing whatever benefits net of costs are captured by the winning city.

Most economists conclude that targeted business-location subsidies, often committing more taxpayer spending per job created than the new employees will earn in a year, do not pay off. State and local politicians may win when they lure Amazon to their city and we undoubtedly will see them standing front and center to take credit for an economic development windfall as they help cut the red ribbon at the shiny new headquarters building.

Don’t be fooled. Such victories impose heavy costs on taxpayers, who stand to lose more than Amazon gains.